NEW YORK (GenomeWeb News) - Many roads lead to Rome, and so do many approaches to a DNA sequence. While several next-generation technologies track the incorporation of labeled nucleotides, others attempt to read DNA as it threads through a nanopore.
Now researchers at Stanford University have demonstrated yet another approach to single-molecule sequencing: They follow a single RNA polymerase enzyme and track its pauses as its travels along the DNA making RNA.
“It is an experiment that was really only a pipe dream many years ago,” said Steven Block, a professor of biological sciences and applied physics at Stanford. “What is truly new about this is we are using the motion — the mechanics of the enzyme — rather than its biochemistry to derive information.”
In their experiment, a proof-of-concept study, the researchers were able to identify correctly 30 out of 32 bases of a target sequence in fewer than 3 minutes of observation time using four DNA molecules.
“This is the smallest amount of DNA that has ever been used to sequence, so this is a record,” said Block, the senior investigator of the research, which appeared in last week’s Science.
Though he says that the speed, accuracy, read length, and throughput of the technique can be improved in principle, it is unclear whether it will have commercial applications, at least in the near term.
“I think it’s a very clever demonstration of what you can do with single molecule techniques, but it’s far too early to say if there are any practical applications of it,” said Stephen Quake, a professor of bioengineering at Stanford University.
Quake is also a co-founder of Helicos BioSciences, which is commercializing his single-molecule sequencing technology. He said his approach, which tracks the incorporation of fluorescent nucleotides, was published in PNAS in 2003 and is the first demonstration of single-molecule sequencing to appear in a scientific journal. Block’s is the second.
Block’s technology relies on two optical traps that each hold one polystyrene bead. One of the beads is attached to a molecule of E. coli RNA polymerase and the other to the end of a DNA molecule. As the enzyme proceeds along the DNA, making RNA, the distance between the two beads changes, which the scientists can measure with angstrom-level precision.
In order to read the bases, the researchers have to run the experiment four times, each time limiting one of the four nucleotides. Whenever the polymerase has to incorporate the limited nucleotide, it pauses for a brief moment. By analyzing the pauses in all four experiments, the scientists can deduce the entire sequence.
The approach has analogies to Sanger sequencing, Block explained: Both require four separate reactions, one for each base, and both need a known primer, or flanking sequence, as a starting point.
The ultimate read length of the technique is limited by the processivity of the enzyme, or the number of bases it will remain on the DNA. In the case of RNA polymerase this is several thousand bases.
Block’s group has already shown that it can follow the motion of the enzyme for more than 2,000 bases with near-base-pair accuracy. “Of course we have not sequenced 2 kilobases. What I am claiming here is that you can find single molecule records which in principle provide base information that is 2,000 base pairs long,” he said.
Complicating read length, though, is the fact that researchers must analyze four records together to obtain a single sequence, and the longer the records become the more likely they are to fall out of alignment, Block said.
However, “there are techniques to get around this issue,” he said, such as using sequences from the middle that were already deciphered for a realignment. “In principle, [read length] can be several thousands of base pairs; in practice, it may be several hundreds of base pairs,” he said.
Reads of several hundred or thousand bases would be an improvement over current next-generation technologies — whose reads range from tens to hundreds — but only “a marginal improvement” compared to Sanger sequencing, according to Quake.
In terms of speed, the technique is limited by how fast the enzyme can proceed when one nucleotide is in limited supply, Block said. Sequencing speed could reach several bases per second and could further increase up to five times if a different enzyme was used, such as a viral RNA polymerase, he said.
Throughput could also be improved by parallelizing the experiment, but that would rely on making the techniques for measuring molecules at the angstrom level more practical, according to Block.
“Right now, we are the only lab in the world that can do this,” he said, adding that “this is a technically challenging experiment.” Just talking loudly in the lab, for example, will disturb the measurement. The next generation of instruments may not rely on optical traps, he said, but on atomic force measurements, for example.
Sequencing accuracy can be improved just by repeating the experiment, he said. “In principle, you could have error rates that approach those of current sequencing techniques by just repeating the assay a half-dozen times,” he said. “Even if you did this a half-dozen times, it would still represent the smallest number of molecules that anyone has ever gotten sequence from,” he added.
According to Quake, researchers prize two parameters in next-generation sequencing technologies developed for re-sequencing and de novo sequencing: throughput and read length.
Improving both of those will be “the major challenges before [Block’s technology] becomes practical in any sense,” he said. “It’s not clearly superior to anything out there in either of the two categories that really matter.”
Block conceded that the technique “is not ready to be put into a $50,000 box in the next few months.” But he said Stanford’s office of technology licensing has already received some inquiries, and he sees commercial potential for his invention in the future.
Why did Block choose E. coli RNA polymerase as the tracking enzyme? “It was an outgrowth of our work on RNA polymerase, on trying to understand fundamental things about how transcription works,” he said.
His lab, a biophysics group focusing on basic research into enzymes involved in transcription, originally developed the instrumentation for studying E. coli RNA polymerase. In 2005, the researchers published a study in which they were able to discern the steps of the enzyme as it moves from one base to the next. “If you can ... actually see an enzyme pass from one base to the next along DNA, then that immediately suggests a way to sequence DNA,” Block said.
Block points out that even though his research was not funded through NHGRI’s Advanced Sequencing Technology program, “we have made more progress towards real sequencing than a lot of these companies have.”
“If you are trying to get novel solutions to problems, you are often better off allowing people to explore possibilities afforded by fundamental research, and not simply request a technology” Block said. “If [NIH] simply just fund[s] applications, they will tend to develop technologies that are currently in hand but they won’t spur the development of new stuff in quite the same way. Nothing spurs the development of new technology more than basic research.”
Julia Karow covers the next-generation genome-sequencing market for GenomeWeb News. E-mail her at [email protected]